HN Simulatornew | past | comments | lists | submit | sgc's commentslogin

Debian testing seems to be more stable than what most distros put in main. I bend the rules and get most everything from backports to have the best of both worlds (stable + testing).

From my experience (albeit 20 years ago) Debian testing was basically unusable since it could get into an unworkable state quite easily. You were much better off running unstable instead if you needed the absolutely most recent versions of stuff. It would still break occasionally but would get fixes would be available much sooner.

Backports are quite variable. Distro packaging systems are really awkward for software from language ecosystems with many dependencies. I end up having to install a static build of podman if I want a recent version. A backport basically seems impossible.


I have been running with backports for years and can't remember the last time I had a problem. Perhaps things have matured over the years to explain the different experiences.

Backports don’t break a stable install, they tend to either work or are incompatible due to underlying library mismatches (eg libc upgrade.)

However running a testing install can result in a broken system after an apt upgrade and it can stay broken for some time while fixes work their way in from unstable.


Me too

Me three. I was surprised how same it all was. Except for the snaps! Debian doesn't force you to use snaps. This was perhaps the top reason I switched.

Ha! Ditto. I abhor snaps.

For ocr, you have to practice, practice, practice with your prompts on small sample sizes before one-shotting a book. I create a json schema for the output that describes the structure in detail (using ai from my prose descriptions of structure of course). I have specific tough pages I use on openrouter to test every new model I am interested in. Gemini has and does absolutely dominate for this. But flash can be better because it doesn't overthink things.


I did do test images, but apparently not enough. It's not 100% of images it drops the second column on either. On many it gets both, on some it only gets the second column.

I also found that it ignores other instructions sporadically ("Don't use letter variants or ligatures in the output, use the normal Danish/Norwegian alphabet when at all possible", still a few Unicode tall s'es in the output.)

It's good, but it's so frustrating since "worse" models manage to follow such instructions much better.


Please stay away from Dell. Posting from the worst laptop I have ever owned, my Dell XPS that cost over 2.5k usd new. Yes, I hate it enough to rage-post about it on HN.

At least it sort of works after using the best thermal compound available, adding thick feet, and using it only on a cutting board to make sure an uneven surface like my lap doesn't reduce flow. Also, never place the charger on the carpet or it will shut itself off from thermal overload. Everything still randomly overheats with minimal workloads, but mainly it sort of works now :)


I've also experienced numerous thermal issues with Dell XPS'es (keyboard de-gluing and popping up, for instance). Hope their laptops are better now, but I wouldn't vouch for them.


FWIW, I've used their Latitude line of laptops and they've been excellent on Linux.

But that could possibly be an exception, and the rest of the company is staffed with shit flinging monkeys, at least that's what I infer from the horrid design of the XPS product page.


My own experience doesn't match yours. I've bought 6 XPS laptops and a couple of desktops (including Alienware) over the last 20 years and have had a pretty good experience with all of them. They always had nice screens, good performance, and held up to the stresses and strains of travel and heavy usage.

That doesn't mean I haven't had a handful of issues, such as several swollen batteries that caused problemss with the touch pad and keyboard. But Dell fixed the ones (onsite!) that occurred during the warranty period and I was able to fix the ones that occurred outside the warranty period.

The only issue that really sticks with me is one I had with motherboard fan controls (not lights, but fan speed) in an Alienware R6 that could only be engaged and adjusted via Windows software. Thanks to proprietary Dell bios shenanigans I couldn't get control of the fan via Linux no matter what I tried.

That said, I don't find this "Create Edition" model appealing for so many of the reasons other posters have pointed out (price, Windows preinstalled, etc.). And I really dislike monitors that use uncommon resolutions (I think this one is 3200x2000).


I've owned several generations of XPS and whenever I need a new laptop XPS is my choice. So far I've loved them. :) Haven't run into any overheating issues so far, but I'm sure we use ours differently.


I had a Dell latitude E6400 for many years, and it was rock solid for me with zero complaints. After that I had an XPS 9550, and it had a lot of issues over just ~3-4 years: issues with power connector, audio stopping working, touchpad issues, etc. Since then I decided to go back to Thinkpad.


I have an xps15 that yes, runs hot when I really use it, but it is otherwise fine. May be a model year situation. I have a 9510 from 2021 and it still suits fine as a daily driver for node / angular / postgres etc running in containers as I do dev work.


My impression has been that Dell generally provides better long term support such as firmware security updates than Asus. Curious what is recommend instead, Lenovo perhaps?


To each their own my dude. I’ve been on XPS13 since 2014, all three have been solid and are still working. Latest one is from 2023. Developer edition with Ubuntu preinstalled.


Every Dell XPS Ive ever had has broken within a year or two. However the warranty is amazing and they fixed it within a few weeks.

I have a Dell Precision that's a tank though.

Maybe QC differences in consumer vs business models.


Seems like 95% of devices could be managed with a hardware filter that only allows through a handful of very rudimentary signals (more like Morse code than a full digital transmission), and then of course a properly designed on-device API that defines those signals and doesn't allow you to trigger a defrost remotely, when that should only ever be done with somebody physically present for example.


Defrost cycles do not require someone to be physically present. These are happening multiple times per day, per evaporate, per cold box. A moderate grocery store or cold storage warehouse will have dozens, it not hundreds, of defrosts per day. They NEED central dispatching and/or remote controls. Controls engineers have already thought of these things. They DO depend on rudimentary signals to schedule these events and coordinate defrosts so that only a selection of evaps defrost at a time and the box can maintain cold temps. It also promotes temperature stability, energy efficiency, and keeps equipment from being overloaded.


Morse code is a fully digital transmission.

How exactly would a more obscure (or even “analog”) encoding help? This seems orthogonal to reducing surface complexity and remotely executable command subsets.


The point is having the digital transmission simple enough that you can use fully *physical* filters to discard everything but the 3-4 messages you need through. Of course it is still digital, but my point is, it is a physically enforced studied, safe subset and doesn't allow for full network access. There are of course a variety of ways to further firewall that, but something needs to act like as much of an air gap as possible while allowing for remote communications.


How do you authenticate such a system? 3-4 “quasi analog” messages seem incredibly easy to replay, if nothing else. If you need to depend on something actually cryptographic to make the system replay proof, you can just as well use that system for all aspects of access control. This makes no sense.


Well it would receive control signals from an on-site relay server with the hardware to communicate with it, and that would of course also be firewalled and use encryption to communicate with the outside world. That server could always in theory become compromised despite your best efforts, so the benefit is obvious: don't create even the theoretic possibility of being hacked by a class of message that should never exist.

The idea is not to solve everything with one simple trick, but to reduce the attack surface as much as possible in hardware. If you are dumb enough to make your 4 messages something like off, on, hard reset, random dangerous maintenance protocol, then it does nothing. If instead it encodes 4 useful and typically used states that can't destroy your system, then you have massively reduced the requirement for onsite, and accepted that more serious issues require onsite control.


Oh, so basically, reduce the vocabulary that can be communicated remotely compared to what's available locally? I can get behind that.

It's not an easy balance to strike, though; the less capable the remote interface is, the less useful it is. "Only completely harmless commands" might simply not be good enough. For example, is "deactivate electricity output of this solar panel" (e.g. because market conditions make it unprofitable to supply at a given instance) harmless or not? What about "increase supplied frequency by 0.01 Hz to stabilize the grid"?


System needs to be carefully designed by experts for sure. Apparently that is a hard ask given the current state of affairs.


Absolutely, what an incredibly low-effort click hijack farm it has become.


Don't do that! v041 has been compromised upstream by a supply chain attack and big fan doesn't want us to know. You have to use v040 until it has been fixed. Also, used the hacked flashing software because the official version is spyware. Did you even read page 723 of the support thread?? I can give you more info when I am done patching my toaster.


Can you give me more information on the toaster? I am tired of it toasting slurs into my unlicensed bread....


Your comment is a bit inaccurate.

1. There are laws already. It is an enforcement problem.

2. The Epstein-type crimes appear to be more prevalent in other branches of government and other parts of society than specifically in the legislative.

3. Enforcement problem is specifically in the executive, of which enforcement is a part.


Saying its an enforcement problem is itself inadequate, because there seem to be nothing really happening and there are endless numbers of powerful and wealthy people who are "in the Epstein files" that gets out in the media, yet somehow none or almost none of them are being prosecuted now or in the past. Some of the now adult victims have made public acknowledgement, congress passed a law saying the details must be disclosed, and the trump admin refuses to do that. Congress is captured by republicans and repubs in congress are captured by trump. So nothing happens.

At the highest level the release of the files is blocked (i.e. congress and the president are the highest level).


>The Epstein-type crimes appear to be more prevalent in other branches of government and other parts of society than specifically in the legislative.

Based on what?


It's a two-edged sword. Compared to other governments your own government when good is the best to promote your good. But when bad, it is by far the worst abuser.


Sorry if this is in that article, but I am on my phone and can't see it. How much would this cost to batch generate an audiobook? Right now I just listen to things in the 11 labs app which is free, but I would rather just generate audio files.


Roughly 5-10$ for 10h, assuming you few-shot it.

Price per hour:

- 3.8 Flash TTS, standard: $0.81

- 3.8 Flash TTS, batch: $0.41

- 3.8 Flash‑Lite TTS, standard: $0.54

- 3.8 Flash‑Lite TTS, batch: $0.27


That is the biggest difference here for me. I have not looked at every solution, but many. They are either much more expensive or garbage quality. For example my next target audiobook is a monster 250k words, 1.5m characters, so about $15 here or $75 using elevenlabs (0.05 per 1k characters). For me that is the difference between I will or I won't use it.


Until the end of the year, then double that. And that's only the audi output though the text input shouldn't cost much in comparison.

Official pricing can be seen here: https://ai.google.dev/gemini-api/docs/pricing#gemini-3.8-fla...


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: